Anshul Chaturvedi profile photo 1 on Tal

Anshul Chaturvedi

Staff Software Engineer @ Google | Distributed Systems · Reliability & Observability · LLMs & Agents in Production

Bengaluru

11+ years of work experience

Google
Nov 20241 yr 10 mosPresent

Bengaluru

Working on Search Platforms — building the infrastructure that keeps Google Search and AI Mode reliable and observable at scale. - Grew the Bangalore Search feature reliability team (7 SWEs), driving full ownership transition from the US org with zero service disruption. - Real-time monitoring that catches user-facing feature failures — processing ~200K events/s across Search and AI Mode, covering 250+ features from 70+ teams. - LLM evaluation frameworks that bring observability to non-deterministic, prompt-driven features where rule-based monitoring isn't as effective. - A RAG based platform that generates semantically complex evaluation queries usually rare in production — used by 20+ AI Mode teams across model training, evals, and prompt tuning.
Google
May 20222 yrs 7 mosNov 2024

Bengaluru

Amazon
Jul 20201 yr 9 mosMar 2022

Bengaluru

Built the ad moderation platform that reviews 3M+ ads daily — worked on on scalability, fairness across clients, and optimisation problems for the most efficient use limited human content reviewer capacity.
InMobi
May 20203 mosJul 2020

Bengaluru, Karnataka, India

InMobi
Sep 20181 yr 8 mosApr 2020

Bengaluru, Karnataka, India

Built high-throughput data infrastructure for InMobi's affiliate marketing platform — real-time events ingestion, analytics pipelines and cloud infrastructure at scale.
FactSet
Nov 20161 yr 10 mosAug 2018

Hyderabad, Telangana, India

FactSet
Jun 20151 yr 5 mosOct 2016

Hyderabad Area, India

Developed market analytics and reporting solutions within the FactSet ecosystem for portfolio managers.
S&P Capital IQ
May 20143 mosJul 2014

Gurgaon, India

Anshul’s AI Stack

aistudio
aistudio
antigravity
antigravity
claude code
claude code
gemini
gemini
geminicli
geminicli
obsidian
obsidian

Education

Visvesvaraya National Institute of Technology

Bachelor of Technology (B.Tech.), Computer Science

Visvesvaraya National Institute of Technology

20114 yrs2015

Join Anshul and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Anshul Chaturvedi on Tal

Headline

Staff Software Engineer @ Google | Distributed Systems · Reliability & Observability · LLMs & Agents in Production

About

I'm a Staff Software Engineer (L6) at Google, working at the intersection of reliability, observability and LLMs/Agents in production. At Google, I lead teams and platforms that keep Google Search and AI Mode — marquee consumer products used by billions — observable and reliable at scale. My work spans high throughput real-time monitoring infrastructure, LLM evaluation frameworks, and agentic systems that power feature development for AI Mode. I believe the hardest unsolved problem in shipping LLM-powered products isn't model quality, it's knowing when they're broken in production. Traditional monitoring and observability breaks down when outputs are non-deterministic and "correct" is subjective. That's the problem space I work in. If you're working on reliability and observability for AI systems, or thinking about what it takes to ship LLM-powered products with confidence, I'd love to connect.

Location

Bengaluru

AI stack

aistudio

antigravity

claude code

gemini

geminicli

obsidian

Work experience

Staff Software Engineer at Google, Nov 2024 - Present · 1 yr 10 mos, Bengaluru, Working on Search Platforms — building the infrastructure that keeps Google Search and AI Mode reliable and observable at scale. - Grew the Bangalore Search feature reliability team (7 SWEs), driving full ownership transition from the US org with zero service disruption. - Real-time monitoring that catches user-facing feature failures — processing ~200K events/s across Search and AI Mode, covering 250+ features from 70+ teams. - LLM evaluation frameworks that bring observability to non-deterministic, prompt-driven features where rule-based monitoring isn't as effective. - A RAG based platform that generates semantically complex evaluation queries usually rare in production — used by 20+ AI Mode teams across model training, evals, and prompt tuning.

Senior Software Engineer at Google, May 2022 - Nov 2024 · 2 yrs 7 mos, Bengaluru

Software Development Engineer II at Amazon, Jul 2020 - Mar 2022 · 1 yr 9 mos, Bengaluru, Built the ad moderation platform that reviews 3M+ ads daily — worked on on scalability, fairness across clients, and optimisation problems for the most efficient use limited human content reviewer capacity.

Software Development Engineer III at InMobi, May 2020 - Jul 2020 · 3 mos, Bengaluru, Karnataka, India

Software Development Engineer II at InMobi, Sep 2018 - Apr 2020 · 1 yr 8 mos, Bengaluru, Karnataka, India, Built high-throughput data infrastructure for InMobi's affiliate marketing platform — real-time events ingestion, analytics pipelines and cloud infrastructure at scale.

Software Engineer ll at FactSet, Nov 2016 - Aug 2018 · 1 yr 10 mos, Hyderabad, Telangana, India

Software Engineer at FactSet, Jun 2015 - Oct 2016 · 1 yr 5 mos, Hyderabad Area, India, Developed market analytics and reporting solutions within the FactSet ecosystem for portfolio managers.

Software Development Intern at S&P Capital IQ, May 2014 - Jul 2014 · 3 mos, Gurgaon, India

Education

Bachelor of Technology (B.Tech.), Computer Science, Visvesvaraya National Institute of Technology